Extend generator script by precision and blending
Add an option for precision and blending to the generate.py script.
Note, with this MR, real_t is no longer supported since it is not uniquely defined, it could be float64 or float32.
The default is set to walberla::float64.
Edited by Michael Zikeli
Merge request reports
Activity
Filter activity
176 189 operator_generation.optimizer.opts_arg_mapping[opt.upper()] 177 190 for opt in spec["optimizations"] 178 191 } 179 type_descriptor = types.hyteg_default_type() # TODO 180 blending = hfg.blending.IdentityMap() # TODO 192 type_descriptor = types.hyteg_type(spec["precision"]) 193 # FIXME not sure if we want to throw an error if the given blending is not existing or use identity, as it is now. -
We should throw an error @ab04unyc
requested review from @thoennes
assigned to @ab04unyc
Resolved by Daniel Bauer- Resolved by Michael Zikeli
Before this MR can be merged, the HFG MR hyteg-form-generator !68 must be merged.
-
Last reply by
Michael Zikeli
- Resolved by Michael Zikeli
added 1 commit
- 8e52144d - Change how and where the default is chosen and set for precison and blending.
added 1 commit
- 59cb62b6 - Update requirements.txt to use current version of the HFG master as requirement.
Resolved by Michael Zikeli- Resolved by Michael Zikeli
- Resolved by Michael Zikeli
-
This MR also fixes Issue #1 (closed).
mentioned in issue #1 (closed)